Mesenchymal stem cell-derived microvesicles confer protection against rheumatoid arthritis-associated interstitial lung disease

Abstract

Background

Rheumatoid arthritis-associated interstitial lung disease (RA-ILD) is a serious condition with few treatment options and a poor outlook. Mesenchymal stem cell-derived microvesicles (MSC-MVs) offer a potential cell-free therapy, but their effectiveness and the role of their miRNA content in RA-ILD are not well understood.

Methods

A mouse model of comorbid RA-ILD was created using collagen-induced arthritis and bleomycin-induced pulmonary fibrosis. Mice received high- or low-dose MSC-MVs via tail vein injection. The treatment effectiveness was evaluated by assessing arthritis severity and pulmonary fibrosis through clinical scores, imaging, histopathology, lung function, and lung coefficients. The expression of fibrotic markers in the lungs was measured using immunohistochemistry and Western blotting. Small RNA sequencing of MSCs-MVs was conducted to explore the underlying mechanism.

Results

Administering MSCs-MVs significantly reduced the severity of arthritis and pulmonary fibrosis. Treated mice had better joint and lung scores, with decreased fibrotic areas and lower levels of fibrotic proteins such as anti-α-smooth muscle actin, collagen I, and fibronectin. Small RNA sequencing revealed that MSC-MVs contain microRNAs (miRNAs) (e.g., miR-148a-3p) that target critical pathways in RA-ILD, such as transforming growth factor-β (TGF-β) and Janus kinase-signal transducer and activator of transcription (JAK-STAT) signaling.

Conclusions

Our research revealed that MSCs-MVs significantly protect against RA-ILD in a comorbid model. This protective effect is likely facilitated by the transfer of specific miRNAs that modulate essential profibrotic and inflammatory pathways. These results underscore the therapeutic potential of MSC-MVs as an innovative cell-free approach for RA-ILD treatment and identify miRNA-mediated pathway regulation as a pivotal mechanism.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s13287-026-05119-w.

Introduction

Rheumatoid arthritis (RA) is a chronic systemic autoimmune disorder characterized by persistent synovitis, progressive joint destruction, and multisystem involvement, potentially resulting in severe disability [1]. Interstitial lung disease (ILD) represents one of the most prevalent and serious extra-articular manifestations of RA, with reported prevalence rates ranging from 10% to 61% [2]. Patients afflicted with RA-ILD encounter a markedly elevated risk of mortality, highlighting the urgent need for effective therapeutic interventions. The management of RA-ILD presents a substantial clinical challenge [3]. Although immunomodulatory therapies are standard for treating RA, their effectiveness in modifying the progression of RA-ILD remains limited. This limitation is underscored by research indicating significantly lower retention rates of biologic and targeted synthetic disease-modifying antirheumatic drugs (b/tsDMARDs) in patients with RA-ILD than in those without ILD [4]. Despite the approval of the antifibrotic agent nintedanib for progressive pulmonary fibrosis, including RA-ILD [1, 2], there remains a critical unmet need for treatments capable of concurrently addressing both the articular and pulmonary pathologies inherent in this complex comorbid condition.

In recent years, mesenchymal stem cell-derived extracellular vesicles (MSC-EVs), including exosomes (EXOs) and microvesicles (MVs), have emerged as promising cell-free therapeutic platforms [5]. Compared with whole mesenchymal stem cells, MSC-EVs offer distinct advantages, such as lower immunogenicity, reduced risks of tumorigenicity and vascular occlusion, and improved storage and handling capabilities [6]. The therapeutic effects of MSC-EVs are mediated primarily through the delivery of bioactive molecules, particularly microRNAs (miRNAs), to recipient cells [6]. The functions of MSC-EVs are multifaceted: (1) Immunomodulation: MSC-EVs can inhibit the release of proinflammatory cytokines such as TNF-α and IL-6 and restore the Treg/Th17 balance, thereby mitigating the progression of arthritis [7, 8]; (2) Antifibrotic effects: MSC-EVs demonstrate efficacy in models of pulmonary fibrosis by suppressing transforming growth factor-β (TGF-β)1 signaling, inhibiting myofibroblast differentiation, and reducing collagen deposition [9]; (3) Targeted delivery: Intravenously administered MSC-EVs exhibit a propensity to home to sites of inflammation, such as inflamed joints [8]. Importantly, the therapeutic potential of MSC-EVs is fundamentally associated with their distinct miRNA cargo [10]. For example, MSC-EXOs have been shown to mitigate RA through the delivery of miR-140-3p, suggesting that miRNAs serve as essential mediators in this process [11]. Nevertheless, the specific role and miRNA profile of MSCs-MVs in the context of the comorbid condition of RA-ILD remain largely unexamined, highlighting a substantial gap in the current body of knowledge.

Building on our previous success with human umbilical cord MSC-MVs in treating bleomycin (BLM)-induced pulmonary fibrosis in mice [12], we hypothesized that these MVs could target both joint inflammation and pulmonary fibrosis in RA-ILD through their miRNA cargo. We constructed a mouse model [13] in which collagen-induced arthritis (CIA) and BLM-induced pulmonary fibrosis were combined. We evaluated the therapeutic effects of MSC-MVs on joint and lung issues and analyzed their miRNA profiles. This study is among the first to explore the therapeutic potential and mechanisms of MSCs-MVs in a preclinical RA-ILD model, offering new insights and a promising treatment strategy for this condition.

Method

Statement: The work has been reported in line with the ARRIVE guidelines 2.0.

Preparation and characterization of MSC-MVs

The human umbilical cord-derived mesenchymal stem cells (hUC-MSCs) used in this study were generous gifts from Chengdu Kangjing Biotechnology Co., Ltd. (Chengdu, China). The collection of umbilical cords and the subsequent isolation of MSCs were conducted in strict accordance with the Declaration of Helsinki. The study and the cell donation process were approved by the Medical Ethics Committee of Wenjiang District People’s Hospital, Chengdu City (Approval No. 2024064; approved on December 29, 2024), and the donors signed informed consent. P4 MSCs were cultured in complete medium supplemented with 10% fetal bovine serum (FBS) until passage 5 (P5). Upon reaching greater than 80% confluence, the medium was removed, and the cells were washed three times with phosphate-buffered saline (PBS). The cells were subsequently transferred to serum-free medium and subjected to a 48-hour starvation period at 37 °C with 5% CO2 to facilitate microvesicle (MV) release. The conditioned medium was collected and subjected to sequential centrifugation at 400 × g for 10 min and 2000 × g for 20 min at 4 °C to eliminate cellular debris. This was followed by ultracentrifugation at 50,000 × g for 2 h at 4 °C to pellet the MVs, which were then resuspended in PBS, aliquoted, and stored at − 80 °C. On the day of use, MVs were thawed and diluted accordingly. For transmission electron microscopy (TEM) analysis, MV suspensions were applied to carbon-coated copper grids for 3 to 5 min, blotted, stained with 2% phosphotungstic acid for 1 to 2 min, air-dried, and subsequently imaged. The size and particle concentration of the MVs were determined using nanoparticle tracking analysis (NTA) with a ZetaView instrument (488 nm). The samples were diluted 500 to 2000 times in ultrapure water to achieve particle concentrations of approximately 10^7 to 10^9 particles/mL, and the data were analyzed using the instrument’s software. Additionally, the levels of the characteristic proteins CD9 and TSG101 were detected and characterized using Western blotting (WB).

Animal experiments

Eight-week-old male DBA mice (19–20 g) were obtained from Beijing Huafukang Bioscience Co., Ltd. This study was approved by the Animal Ethics Committee of West China Hospital, Sichuan University (Approval No: 20240515007). The mice were kept under pathogen-free conditions at 22 °C with a 12-h light/dark cycle and had unlimited access to food and water. For the CIA model, on day 0, 100 µL of an emulsion (50 µL Complete Freund’s Adjuvant (CFA; Chondrex, USA) and 50 µL Chicken Type II Collagen (Chondrex, USA)) was injected subcutaneously at the tail base. On day 21, a booster of 100 µL of emulsion (50 µL of incomplete Freund’s Adjuvant (IFA; Chondrex, USA) and 50 µL of Chicken Type II Collagen) was given. To construct the CIA model, a CIA-BLM model was developed in which BLM at a dosage of 3 mg/kg was administered intratracheally to anesthetized mice on day 28. During this procedure, the mice were anesthetized using an animal anesthesia system with 4–5% isoflurane for induction and 1.5–2% isoflurane for maintenance, with medical air as the carrier gas at a flow rate of 1 L/min. Twenty mice were divided into four groups: control (CTL), RA-ILD (CIA + BLM), high-dose MSC-MVs (MVs-high), and low-dose MSC-MVs (MVs-low), with five mice each. Starting on day 29, the MV-high and MV-low groups received tail vein injections of 5 × 10⁸ or 1 × 10⁸ MSC-MV particles, respectively, twice weekly for three weeks. The CTL and CIA + BLM groups received PBS instead. On day 49, all the mice were euthanized. The mice were first deeply anesthetized with 5% isoflurane for approximately 5 min until the disappearance of the pedal reflex, followed by cervical dislocation to ensure clinical death. Lung, knee, and ankle tissues were collected for analysis. The experimental timeline is shown in Fig. 1A.

Fig. 1.

Evaluation of the effects of MSC-MVs on arthritis and joint degeneration in RA-ILD mice with rheumatoid interstitial lung disease (N = 5). A Diagram of the experimental scheme. B Representative images of hind paws from each experimental group. C Clinical arthritis scoring chart for the respective groups of mice. D Representative three-dimensional reconstructed microcomputed tomography (micro-CT) images of knee joints, with arrows indicating areas of interest. E–I Quantitative analysis of bone microarchitecture parameters, including E bone volume fraction (BV/TV), F bone surface density (BS/TV), G trabecular number (Tb.N), H trabecular thickness (Tb.Th), and I trabecular separation (Tb.Sp). J Representative micrographs of H&E-stained knee joint sections (scale bar: 200 μm). K Representative micrographs of toluidine blue (TBO)-stained knee joint sections (scale bar: 200 μm). L–O Semiquantitative histopathological scoring charts for joint sections, detailing L inflammatory infiltration and M synovial hyperplasia scores and N bone erosion scores. M Mankin scoring chart for cartilage assessment using TBO staining. The data are shown as the means ± SEMs. ∗ indicates a statistically significant difference between any two groups. ∗p < 0.05; ∗∗p < 0.01; *** p < 0.001. RA-ILD rheumatoid arthritis-associated interstitial lung disease, CII collagen type II, CFA complete Freund’s adjuvant, IFA incomplete Freund’s adjuvant, BLM bleomycin, BV/TV bone volume fraction, BS/TV bone surface density, Tb.N trabecular number, Tb.Th trabecular thickness, Tb.Sp trabecular separation, H&E hematoxylin and eosin, TBO toluidine blue O, CTL control, CIA collagen-induced arthritis, SEM standard error of the mean

Arthritis score

Arthritis severity was assessed by visually inspecting swelling and inflammation in all four paws. Clinical scoring began on day 28 and occurred biweekly. Each paw was graded on a 0–4 scale: 0 for no swelling/erythema, 1 for mild swelling/erythema in small toe joints, 2 for moderate swelling up to metatarsal bones, 3 for severe swelling below the ankle, and 4 for maximum swelling/erythema, including the ankle with joint ankylosis. The total arthritis score for each mouse, from 0 to 16, was the sum of the scores from all the paws.

Assessment of pulmonary function by whole-body plethysmography (WBP)

After the system was calibrated, the mice were placed in the restraint-free chamber of the WBP system, where their respiratory data were recorded for 5 min of calm breathing. The plethysmography signals were analyzed to determine standard respiratory parameters, including respiratory frequency (f, breaths/min), tidal volume (VT, mL), minute ventilation (MV, mL/min), peak inspiratory flow (PIF, mL/s), peak expiratory flow (PEF, mL/s), mid-expiratory flow (EF50, mL/s), expiratory time (Te, s), inspiratory time (Ti, s), relaxation time (Tr, s), and airway resistance, as indexed by enhanced pause (Penh).

Lung coefficient calculation

Lung tissues were carefully collected and cleared of connective tissue during dissection. After blotting on filter paper to remove fluids and blood, the samples were photographed. The lung wet weight was measured, and the lung coefficient was calculated as the lung wet weight (mg) divided by the body weight (g), in mg/g.

Micro-CT analysis of the knee joint

After dissection, the left hind limb segments (femur, knee joint, and tibia) were cleaned of skin and muscle, fixed in 4% paraformaldehyde for 48 h, and then rinsed and scanned using a NEMO^® Micro-CT system (model NMC-200; PINGSENG Healthcare (Kunshan) Inc., China). The scanning settings were 0.05 mA current, 90 kV voltage, 30 mm FOV, and 25 μm voxel size. 3D reconstruction was performed with Avatar software. For analysis, a 2 mm region from the tibial metaphysis growth plate was examined for bone tissue, and trabecular parameters such as bone volume fraction (BV/TV), bone surface density (BS/TV), trabecular number (Tb.N), trabecular thickness (Tb.Th), and trabecular separation (Tb.Sp) were calculated.

Histological examinations

After euthanasia, lung tissues and knee joints were quickly dissected and fixed in 10% formalin for 24 h. Lung tissues were processed into 5 μm paraffin sections, which were stained with hematoxylin and eosin (H&E) or Masson’s Trichrome (Servicebio, China). The Ashcroft score was used to assess pulmonary fibrosis severity on a 0‒8 scale in these Sect [14]. For immunohistochemistry, the sections were treated to block endogenous peroxidase activity (Servicebio, CA), incubated with an anti-α-smooth muscle actin (α-SMA) primary antibody (Abcam, UK), incubated with an HRP-conjugated secondary antibody, and developed with DAB. Finally, the slides were counterstained, dehydrated, cleared, and mounted.

After fixation and decalcification, the mouse knee joint samples were dehydrated in ethanol, cleared in xylene, and embedded in paraffin. The tissues were sectioned at 5 μm, deparaffinized, rehydrated, and stained with H&E for histopathological evaluation and toluidine blue O (TBO; BIOSSCI, China) for cartilage matrix assessment. Blinded observers scored the H&E-stained sections for bone erosion, inflammatory infiltration, and synovial hyperplasia (0–3 scale). The integrity of the TBO sections was evaluated using a modified Mankin score (0–3) for proteoglycan loss and structural changes.

Western blotting

Protein was extracted from animal lung tissues using cell lysis buffer (Beyotime, CA) supplemented with protease and phosphatase inhibitors (1:100; Beyotime, CA). The protein concentration was determined with a BCA assay (Beyotime, CA). Equal amounts of protein were heated with SDS–PAGE buffer, and 25 µg of protein was separated on 10% polyacrylamide gels (Bio-Rad, USA) and then transferred to PVDF membranes. The membranes were blocked with 5% BSA (Servicebio, CA), incubated overnight with anti-Col Ⅰ (1:2000; Abcam, USA), anti-fibronectin (FN) (1:5000; Abcam, USA), and anti-β-antin (1:10000; TransGen Biotech, China) primary antibodies and then with a secondary antibody (1:10000; ZEN-BioScience, CA). Bands were detected and analyzed using Image Lab software.

Small RNA sequencing and bioinformatic analysis

RNA quality in the MSC-MVs was assessed with an Agilent 2100 Bioanalyzer. Libraries were prepared using the Small RNA Library Kit for Illumina and sequenced on the Illumina NovaSeq 6000 platform. After quality control, the miRNAs were identified and annotated via the miRBase database, and family analysis was performed using miFam.dat. Targets for the top 10 miRNAs were extracted from miRTarBase and TarBase, with a focus on interactions confirmed by evidence from a reporter assay. The target lists were merged, duplicates removed and subjected to Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analyses using the R package clusterProfiler (v4.4.4). A core regulatory network of “miRNA–target gene–pathway–disease” interactions was then created with Cytoscape.

Statistical methods

Data analysis involved one-way ANOVA with the Sidak post hoc test for multiple comparisons using GraphPad Prism 8.0.1. The Mann‒Whitney U test was used for nonnormally distributed data between two groups. The results are shown as the mean ± SEM for normal data and median for nonnormal data. A p value of < 0.05 was considered to indicate statistical significance.

Results

MSC-MVs alleviated arthritic symptoms and joint destruction in mice with RA-ILD

MSC-MVs were isolated and characterized using TEM, NTA, and WB analysis. MSC-MVs exhibited a typical cup-shaped morphology (Figure S1 A) and Median (X50) of diameter at 151.1 nm (Figure S1 B). MSC-MVs expressed the CD9 and TSG101 proteins (Figure S1 C).

To evaluate the therapeutic potential of MSC-MVs in RA-ILD, we administered them intravenously to mice with this disease. As shown in Fig. 1A, MSC-MVs significantly reduced joint swelling and clinical arthritis scores in the mice, regardless of dose (P < 0.05; Fig. 1B, C). Micro-CT analysis confirmed improvements in articular surfaces and cartilage erosion (Fig. 1D). Additionally, MSC-MVs significantly reversed bone microarchitecture deterioration (BV/TV, BS/TV, Tb.N, Tb.Th, Tb.Sp) at both doses (P < 0.05; Fig. 1E-I). Histopathological analysis with H&E staining confirmed that MSC-MVs reduced joint space narrowing, synovial hyperplasia, inflammatory infiltration, and chondrocyte loss (Fig. 1J, K). Semiquantitative scoring indicated significant improvements in the treated groups (P < 0.05; Fig. 1L, M). The protective effect against cartilage degradation was also supported by TBO staining and a lower Mankin score (P < 0.05; Fig. 1O). These findings indicate that MSC-MVs significantly protect against joint lesions in a mouse model of RA-ILD by reducing synovial inflammation, preserving cartilage integrity, and enhancing bone microstructure.

MSC-MVs attenuated pulmonary fibrosis and improved lung function in RA-ILD mice

Assessment at the endpoint revealed that MSC-MVs clearly affected the lung condition of mice with RA-ILD. The model group had increased lung tissue volume and red coloration, indicating edema and congestion (Fig. 2A). Compared with that in the control group, the lung coefficient in the model group significantly increased, while that in the MSC-MV treatment group recovered (p < 0.05), with no significant difference between the two dosage groups (Fig. 2B).

Fig. 2.

Evaluation of the therapeutic effects of MSC-MVs on lung tissue and function in RA-ILD mice (N = 5). A Representative macroscopic images of lung tissues from each group. B Statistical chart of the lung coefficient for each group. C–L Results of lung function parameter measurements, including C Te, D Ti, E TR, F F, G Vt, H Mv, I PEF, J PIF, K EF50, and L Penh. M, N Representative micrographs of H&E-stained lung tissue sections. Scale bars: 1 mm (M); 100 μm (N). O, P: Representative micrographs of Masson’s trichrome-stained lung tissue sections. Scale bars: 1 mm (O); 100 μm (P). Q Statistical chart of Ashcroft fibrosis scores for lung tissues. The data are shown as the means ± SEMs. ∗ indicates a statistically significant difference between any two groups. ∗p < 0.05; ∗∗p < 0.01; *** p < 0.001. RA-ILD rheumatoid arthritis-associated interstitial lung disease, H&E hematoxylin and eosin, Te expiratory time, Ti inspiratory time, Tr relaxation time, F respiratory frequency, Vt tidal volume, Mv minute ventilation, PIF peak inspiratory flow, PEF peak expiratory flow, EF50 mid-expiratory flow, Penh enhanced pause, SEM standard error of the mean

Lung function tests revealed crucial findings: the model group showed prolonged Ti, which significantly decreased after MSC-MV treatment (P < 0.05), and both high and low doses were effective, with no significant difference between them (Fig. 2C). The Ti and Tr in the model group decreased after treatment (Fig. 2D, E). Additionally, the f, Vt, Mv, PIF, PEF, and EF50 of the model group decreased, and all the values partially recovered after treatment (Fig. 2G-K). Penh expression was significantly greater in the model group but decreased markedly with high-dose MSC-MV treatment (P < 0.05), and the level in the low-dose group also tended to decrease (Fig. 2L).

Histological analysis revealed that MSC-MV treatment significantly improved lung structure, reduced inflammation, and decreased collagen deposition in the model group, as evidenced by H&E and Masson’s trichrome staining (Fig. 2M-P). The Ashcroft fibrosis score, which was elevated in the model group, significantly decreased after treatment (P < 0.05) (Fig. 2Q).

These findings collectively indicate that MSC-MVs significantly mitigate pulmonary fibrosis in RA-ILD mice by improving lung histoarchitecture, decreasing collagen deposition, and ameliorating pulmonary function impairment.

MSC-MVs attenuate extracellular matrix deposition in the lung tissues of mice with RA-ILD

To elucidate the therapeutic mechanisms of MSC-MVs in the pulmonary context of RA-ILD mice, we evaluated the presence of myofibroblasts, which are identified by α-SMA positivity and are pivotal in the pathogenesis of pulmonary fibrosis. Immunohistochemical staining revealed that the model group displayed pronounced increases in α-SMA expression levels (AOD; Fig. 3A, B), the percentage of α-SMA-positive myofibroblasts (Fig. 3C), and their area density (Fig. 3D). Corresponding to these findings, MSC-MV treatment at two different doses resulted in a significant reduction in all three parameters (Fig. 3A–D). Further analysis of the extracellular matrix components of fibrotic tissue, specifically Col I and FN, was performed. Compared with that in the control group, Col I expression in both the MSC-MV treatment groups markedly increased but significantly decreased in the model group (Fig. 3E, G). FN expression exhibited a similar pattern of alteration (Fig. 3E, F). Collectively, these findings indicate that MSC-MVs treatment effectively impedes critical processes involved in pulmonary fibrosis, including myofibroblast activation and excessive deposition of the extracellular matrix.

Fig. 3.

Effects of MSC-MVs on extracellular matrix deposition in the lung tissues of RA-ILD mice (N = 5). A Representative micrographs of α-SMA immunohistochemical staining in lung tissue sections. Scale bars: 200 μm (upper); 50 μm (lower). B–D: Quantitative analysis of α-SMA expression: B AOD of α-SMA, C percentage of α-SMA-positive a cells, and D number of α-SMA-positive cells per unit area. E Representative Western blot analysis of Col I and fibronectin FN protein expression. The corresponding uncropped full-length blots are included in Figure S2 (n = 4). F, G Statistical charts of the quantitative analysis of Col I and FN protein expression levels. The data are shown as the means ± SEMs. ∗ indicates a statistically significant difference between any two groups. ∗p < 0.05; ∗∗p < 0.01; *** p < 0.001. α-SMA alpha-smooth muscle actin, Col I Collagen I, FN Fibronectin, SEM standard error of the mean, AOD average optical density

miRNA cargo composition of MSC-MVs and functional prediction reveals potential mechanisms against RA-ILD

To elucidate the molecular mechanisms underlying the therapeutic effects of MSC-MVs, we conducted a sequencing analysis of their miRNA cargo. The findings indicated that MSC-MVs were enriched with a distinct set of miRNAs, including hsa-miR-21-5p, hsa-miR-148a-3p, hsa-miR-222-3p, hsa-miR-151a-3p, hsa-miR-100-5p, hsa-let-7i-5p, and hsa-miR-26a-5p (Fig. 4A). These miRNAs exhibited consistent expression patterns across three independently prepared batches of MSC-MVs, demonstrating reproducible enrichment (Figure S3).

Fig. 4.

Analysis of miRNA Profiles in MSC-MVs and Their Predicted Regulatory Networks in RA-ILD. A Compilation of the top 10 most enriched miRNAs in MSC-MVs. B Bubble plot of the results of the GO and KEGG enrichment analyses of the target genes of the top miRNAs in the MSC-MVs. Bubble size represents the number of target genes enriched in the corresponding term or pathway. Bubble color represents the significance level of the enrichment (-log₁₀(adjusted P value)), with the gradient from blue to red indicating increasing statistical significance. C Central “miRNA–target gene–pathway–disease” regulatory network demonstrating how key miRNAs coordinately modulate RA-ILD through multiple pathways. GO Gene Ontology, KEGG Kyoto Encyclopedia of Genes and Genomes, MSC-MVs mesenchymal stem cell-derived microvesicles, RA-ILD rheumatoid arthritis-associated interstitial lung disease

Utilizing the ten most abundant miRNAs, we retrieved their target genes from databases containing experimentally validated miRNA–target interactions, revealing an extensive regulatory network. Pathway enrichment analysis indicated that these miRNAs were significantly associated with key signaling pathways relevant to RA-ILD. Importantly, all the core miRNAs converge on the phosphatidylinositol 3-kinase-akt (PI3K-Akt) signaling pathway by targeting various components of this pathway.

Subsequent analysis revealed multiple hub miRNAs with the capacity to concurrently target pathways related to both arthritis and pulmonary fibrosis. Notably, miR-21-5p was found to target transforming growth factor beta receptor 2 (TGFBR2), SMAD family member 7 (SMAD7), and phosphatase and tensin homolog (PTEN) within pathways associated with pulmonary fibrosis, as well as interleukin 1 beta (IL1B), programmed cell death 4 (PDCD4), and MMP9 within pathways related to arthritis. Additionally, miR-148a-3p was shown to target signal transducer and activator of transcription 3 (STAT3) in arthritis-associated pathways and SMAD2 and Rho-associated coiled-coil containing protein kinase 1 (ROCK1) in pathways related to pulmonary fibrosis. At the level of shared targets, several miRNAs, including hsa-miR-5p and hsa-let-7a-5p, were predicted to target IL-6, while insulin-like growth factor 1 receptor (IGF1R) was predicted to be targeted by several miRNAs, including hsa-miR-21-5p, hsa-miR-100-5p, and hsa-miR-148a-3p. Both IL6 and IGF1R are involved in fibrotic and inflammatory pathways.

Disease-specific pathway analysis revealed that in pathways associated with rheumatoid arthritis, these miRNAs target critical molecules within the Janus kinase-signal transducer and activator of transcription (JAK-STAT) signaling pathway, specifically STAT3, STAT5A, and IL6. These molecules are central to the IL-17 and TNF signaling pathways and target key elements such as IL1Bβ, IL6, and the Fos proto-oncogene, AP-1 transcription factor subunit (FOS), as well as genes involved in leukocyte migration. In pathways related to pulmonary fibrosis, hsa-miR-21-5p, hsa-miR-148a-3p, and hsa-miR-26a-5p collectively target essential components of the TGF-β signaling pathway, including TGFBR2, SMAD2, and SMAD4. Additionally, hsa-miR-148a-3p targets the MET proto-oncogene, receptor tyrosine kinase (MET), integrin subunit alpha 5 (ITGA5), and ROCK1 in the focal adhesion signaling pathway. Moreover, these miRNAs target multiple genes involved in extracellular matrix organization, such as MMP9 and IL6, as well as those involved in epithelial‒mesenchymal transition (EMT), including the enhancer of zeste 2 polycomb repressive complex 2 subunit (EZH2).

In conclusion, our bioinformatics analysis revealed that the specific miRNAs encapsulated within MSC-MVs have the potential to collectively target a complex gene network. This network includes pathways common to RA-ILD, such as the PI3K-Akt pathway, and serves to interlink the molecular regulatory frameworks of arthritis and pulmonary fibrosis. This interconnection is facilitated through central miRNAs, such as miR-21-5p and miR-148a-3p, and shared target genes, including IL6 and IGF1R.

Discussion

This study presents experimental evidence demonstrating that MSC-MVs concurrently improve both joint and lung pathologies in a complex model of RA-ILD. Our findings are consistent with the previously reported therapeutic potential of MSC-EVs in the context of inflammatory and fibrotic diseases [15, 16]. Specifically, MSC-MVs significantly reduced the severity of arthritis and preserved joint architecture, corroborating the established immunomodulatory role of MSC-EVs in restoring the balance between Tregs and Th17 cells and in mitigating the pathogenic effects of fibroblast-like synoviocytes (FLSs) in RA models [8, 17, 18]. The observed reduction in cartilage degradation, maintenance of collagen integrity, and reversal of bone microstructural damage aligns with the complex and multifaceted mechanisms through which MSC-EVs target joint destruction.

The protective effects of MSC-MVs are primarily initiated through the attenuation of systemic inflammation. Empirical evidence suggests that MSC-EVs can markedly suppress the secretion of critical proinflammatory cytokines, such as IL-1β, TNF-α, and IL-6, from peripheral blood mononuclear cells [7]. This suppression mitigates the inflammatory environment that contributes to cartilage and bone degradation. Moreover, MSC-EVs play a significant role in immunomodulatory and “osteoimmunology” networks. They facilitate the reestablishment of immune homeostasis, as exemplified by mechanisms such as the ability of miR-148a-3p in gingival MSC-EVs to modulate IKKB/nuclear factor-kappa B (NF-κB) signaling to restore the Treg/Th17 balance [8] and the ability of bone marrow MSC-Exos to increase PR domain zinc finger protein 1 (PRDM1) expression in CD4 + T cells to inhibit Th17 differentiation [19]. Additionally, this immunomodulatory capacity is intrinsically connected to the preservation of bone and cartilage integrity. For example, Exos containing miR-150-5p have been shown to concurrently inhibit osteoclast differentiation and enhance Treg cell function through the aryl hydrocarbon receptor (AhR)/cytochrome P450 family 1 subfamily A member 1 (CYP1A1) signaling pathway [20]. This provides an “osteoimmunology” framework for understanding the observed reduction in bone erosion and preservation of cartilage integrity. Furthermore, the observed suppression of inflammatory infiltration is consistent with previous findings that interferon-beta (IFN-β)-preconditioned MSC-EVs more effectively inhibit the migration and activation of RA FLS [15]. Additionally, miR-34a-enriched EVs have been reported to induce apoptosis in RA FLS via the cyclin I/ataxia telangiectasia mutated (ATM)/ATM and Rad3-related (ATR)/tumor protein p53 (p53) pathway [21]. Importantly, the nondose-dependent efficacy of MSC-MVs is reminiscent of findings in collagen-induced arthritis models, where even low doses achieved maximal therapeutic effects, potentially owing to the rapid saturation of signaling pathways [22].

In pulmonary tissues, MSC-MVs exhibited significant antifibrotic effects, as indicated by decreased collagen deposition, increased Ashcroft scores, and reduced aggregation of α-SMA-positive myofibroblasts. These observations align closely with the established mechanisms of MSC-EVs in models of lung disease. Specifically, MSC-EVs directly inhibit migration, collagen gel contraction, and the expression of α-SMA and ECM components in lung fibroblasts stimulated by TGF-β1 through integrin alpha V-mediated focal adhesion kinase (FAK)/mitogen-activated protein kinase (MAPK) signaling pathways [23]. In addition to modulating fibroblast activity, MSC-EVs directly protect alveolar epithelial cells by increasing the expression of glycosaminoglycans and surfactant proteins, thereby maintaining microenvironmental homeostasis [24], and they also attenuate radiation-induced DNA damage via the ATM/p53/p21 signaling pathway [25]. The observed decrease in inflammatory infiltration is consistent with previous findings that MSC-EVs downregulate oxidative stress regulators, such as nuclear factor erythroid 2–related factor 2 (Nrf2) and heme oxygenase-1 (HO-1), as well as key inflammatory mediators, including NF-κB p65 and IL-1β [26]. Recent studies have further demonstrated that MSC-EVs derived from human umbilical cords mitigate silica-induced pulmonary fibrosis through the circPWWP2A/microRNA-223-3p/NLRP3 axis [27]. Moreover, these substantial preclinical mechanistic insights have translated into promising therapeutic outcomes: nebulized human umbilical cord MSC-EVs significantly improved survival rates from 20% to 80% in a bleomycin-induced model and showed potential for enhancing lung function in patients [28]. Collectively, these findings underscore the promising clinical translation prospects of our validated MSC-EV therapy.

miRNAs are a phylogenetically conserved class of noncoding RNAs, typically 19–22 nucleotides in length, that play crucial roles in regulating cellular processes such as proliferation, differentiation, and apoptosis [29]. These miRNAs are integral effector molecules within MSC-EVs. MSC-EVs facilitate the transfer of miRNAs and other bioactive molecules to recipient cells through mechanisms such as endocytosis and membrane fusion, thereby influencing the function of target cells [30]. In this study, MSC-MVs were enriched in a specific set of miRNAs, including hsa-miR-21-5p, hsa-miR-148a-3p, hsa-miR-222-3p, hsa-miR-151a-3p, hsa-miR-100-5p, hsa-let-7i-5p, and hsa-miR-26a-5p. These miRNAs are key mediators of the therapeutic effects observed in the MSC-MVs in the RA-ILD model. A comprehensive analysis revealed that this combination of miRNAs targets a complex gene network, converging on critical signaling pathways implicated in the progression of RA-ILD.

The therapeutic potential of this miRNA combination lies in its synergistic regulation of pathways critical to both RA and ILD. Notably, all the identified miRNAs targeted the PI3K-Akt signaling pathway through various genes. This pathway is a central hub for cell survival, proliferation, and metabolism [31–33], and its aberrant activation is closely linked to synovial hyperplasia in RA [34] and fibroblast activation in ILD [35]. Concurrent delivery of multiple miRNAs targeting this pathway from different angles may yield more potent suppression than targeting a single node does.

Hsa-miR-21-5p is of particular interest because of its extensive involvement in fibrosis and inflammation. Our network analysis revealed that this miRNA targets pivotal nodes in both RA, including PTEN [36] and PDCD4 [37], and ILD, such as TGFBR2 [38] and SMAD7 [39] Notably, PDCD4 is downregulated in RA [40], and PTEN levels are decreased in both RA [41] and ILD [42], which may limit the role of PDCD4 as a primary mediator of the pro-RA effects of miR-21-5p. Additionally, miR-21-5p targets the central proinflammatory factors IL-1β [43, 44] and MMP9 [45], contributing to the alleviation of joint inflammation and the prevention of chondrocyte degradation. Although miR-21-5p also targets SMAD7, a negative regulator of the TGF-β pathway whose expression is reduced in ILD, its potential to enhance TGF-β signaling may be minimal [46]. Importantly, miR-21-5p further targets critical components of the TGF-β pathway itself, including TGFBR2 [38] and TGFBI [47]. When codelivered with other microRNAs that target essential TGF-β components, such as hsa-miR-148a-3p, which targets SMAD2 [48], and hsa-miR-26a-5p, which targets SMAD4 [49], miR-21-5p establishes a balanced, multilayered regulatory network that finely tunes this crucial signaling pathway.

The therapeutic efficacy of MSC-MVs is likely attributable to their ability to simultaneously modulate multiple disease-specific pathways. In the context of RA, the combination of miRNAs within these MVs may orchestrate robust anti-inflammatory and immunomodulatory responses. This is potentially achieved through direct targeting of the JAK-STAT signaling pathway, including components such as STAT3 [45] and IL6 [51, 52], as well as the inhibition of critical mediators of the IL-17 and TNF signaling pathways, such as ILβ [43, 44], IL6 [51, 52], and FOS [50]. These actions collectively contribute to a reduction in proinflammatory cytokine production and Th17 cell activity [51, 52]. Additionally, the targeting of ADAM17 [53] by hsa-miR-26a-5p may impede leukocyte migration [54], whereas the regulation of SOCS1 [55] and FOS [50] by hsa-miR-222-3p could influence osteoclast differentiation, thereby addressing the bone destruction characteristics of RA [56, 57]. In the case of pulmonary fibrosis, these miRNAs collectively target fundamental fibrotic pathways. Specifically, hsa-miR-21-5p, hsa-miR-148a-3p, and hsa-miR-26a-5p act synergistically to inhibit TGF-β signaling. Furthermore, hsa-miR-148a-3p targets MET [58], ITGA5 [59], and ROCK1 [60], thereby disrupting focal adhesion signaling and interfering with profibrotic cell–extracellular matrix communication [61].

The elevated prevalence of ILD among patients with RA implies the existence of shared pathological mechanisms, commonly referred to as the “joint–lung axis”. Our study provides a molecular foundation for this clinical phenomenon. Network analysis demonstrated that identical miRNAs delivered by MSC-MVs can concurrently target pathways dominated by inflammation in RA and pathways dominated by fibrosis in ILD. For example, hsa-miR-148a-3p targets STAT3 to modulate the JAK-STAT signaling pathway in joints [62], whereas in the lungs, it targets SMAD2 [48] and ROCK1 [60], thereby inhibiting both the TGF-β and focal adhesion pathways. Similarly, IL6, a pivotal node targeted by both hsa-miR-26a-5p and hsa-let-7a-5p, is involved in JAK-STAT signaling in RA and ECM organization in ILD [63, 64]. This multifaceted targeting of shared and interconnected pathways by a single therapeutic agent offers a compelling and sophisticated rationale for the observed synergistic efficacy, reinforcing the potential of MSC-MVs as an optimal strategy for addressing RA-ILD comorbidity.

In conclusion, our findings substantiate a model in which MSC-MVs operate as an intrinsic, multitarget therapeutic delivery system. The miRNA cargo within these vesicles—including hsa-miR-21-5p, hsa-miR-148a-3p, hsa-miR-222-3p, hsa-let-7i-5p, and hsa-miR-26a-5p—functions synergistically to disrupt a disease-promoting gene network that encompasses critical pathways. This coordinated intervention against the inflammatory circuits associated with RA and the fibrotic mechanisms of ILD, including their shared signaling hubs, offers a compelling mechanistic rationale for the observed efficacy in both joint and lung contexts. Our study not only elucidates the potential mechanisms underlying the action of MSC-MVs but also highlights the extensive potential of leveraging endogenous miRNA networks for the treatment of complex, multiorgan diseases. Future research should aim to validate these interactions in vivo and evaluate the therapeutic efficacy of engineered MV-miRNA formulations.

Our findings underscore the importance of miRNA-mediated pathways; however, the specific roles of individual miRNAs necessitate further validation via knockout studies. Additionally, refining the administration protocols and enhancing the scalability of MSC-MVs are essential steps for successful clinical translation.

In conclusion, the use of MSC-MVs is a promising cell-free therapeutic approach for treating RA-ILD through the systemic delivery of regulatory miRNAs to simultaneously mitigate arthritis and pulmonary fibrosis. This “network reset” mechanism, which targets shared nodes within the fibrosis-inflammation cascade, serves as a foundational model for the development of combinatorial nanotherapies aimed at addressing multiorgan comorbidities.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

RA-ILD

Rheumatoid arthritis-associated interstitial lung disease

MSCs-MVs

Mesenchymal stem cell-derived microvesicles

CIA

Collagen-induced arthritis

BLM

Bleomycin

α-SMA

α-smooth muscle actin

Col I

Collagen I

FN

Fibronectin

TGF-β

Transforming growth factor-β

JAK-STAT

Janus kinase-signal transducer and activator of transcription

HR

Hazard ratio

EXOs

Exosomes

DMARDs

Disease-modifying antirheumatic drugs

MSC-EVs

Mesenchymal stem cells-derived extracellular vesicles

miRNAs

MicroRNAs

Treg/Th17

Regulatory T cells and T helper 17 cells

TNF-α

Tumor necrosis factor-alpha

IL-6

Interleukin-6

MMP-9

Matrix metalloproteinase-9

TIMP-1

Tissue inhibitor of metalloproteinase-1

WBP

Whole-body plethysmography

f

Respiratory frequency

VT

Tidal volume

MV

Minute ventilation

PIF

Peak inspiratory flow

PEF

Peak expiratory flow

EF50

Mid-expiratory flow

Te

Expiratory time

Ti

Inspiratory time

Tr

Relaxation time

Penh

Enhanced pause

Micro-CT

Microcomputed tomography

FOV

Field of view

ROI

Region of interest

H&E

Hematoxylin and eosin

TBS

Tris-buffered saline

BSA

Bovine serum albumin

TBO

Toluidine Blue O

GO

Gene Ontology

KEGG

Kyoto Encyclopedia of Genes and Genomes

BV/TV

Bone volume fraction

BS/TV

Bone surface density

Tb.N

Trabecular number

Tb.Th

Trabecular thickness

Tb.Sp

Trabecular separation

PI3K-Akt

Phosphatidylinositol 3-Kinase-Akt

TGFBR2

Transforming growth factor beta receptor 2

SMAD7

SMAD family member 7

PTEN

Phosphatase and tensin homolog

IL1B

Interleukin 1 beta

PDCD4

Programmed cell death 4

STAT3

Signal transducer and activator of transcription 3

ROCK1

Rho-associated coiled-coil containing protein kinase 1

IGF1R

Insulin-like growth factor 1 receptor

IL-17

Interleukin-17; FOS: Fos proto-oncogene

AP-1

Transcription factor subunit

ITGA5

Integrin subunit alpha 5

MET

MET proto-oncogene, receptor tyrosine kinase

EMT

Epithelial–mesenchymal transition

EZH2

Zeste 2 polycomb repressive complex 2 subunit

FLS

Fibroblast-like synoviocytes

PRDM1

PR domain zinc finger protein 1

NF-κB

Nuclear factor-kappa B

AhR

Aryl hydrocarbon receptor

CYP1A1

Cytochrome P450 family 1 subfamily A member 1

IFN-β

Interferon-beta

ATM

Ataxia telangiectasia mutated

ATR

ATM and Rad3-related

p53

Tumor protein p53

FAK

Focal Adhesion Kinase

MAPK

Mitogen-activated protein kinase

Nrf2

Nuclear factor erythroid 2–related factor 2

HO-1

Heme oxygenase-1

Author contributions

LY and LYB conceived, designed and guided the study. LXP, LYH, TZY, ZQP, TCY, WT, HDY, WYL, CL completed the experiments. LXP and LYH analyzed the data. All the authors drafted and revised the manuscript. Xiuping Liang and Yanhong Li contributed equally to this work.

Funding

This work was supported by the Postdoctoral Fellowship Program (Grade C) of the China Postdoctoral Science Foundation (G2C20241135), the Postdoctor Research Fund of West China Hospital, Sichuan University (2025HXBH100), and funding from Grant No. 008420415031.

Data availability

The fastq files were deposited into the NCBI Sequence Read Archive (SRA) database (accession number: PRJNA1397372), https://dataview.ncbi.nlm.nih.gov/object/PRJNA1397372?reviewer=kv4vjh7u9fkboa01ne8veo4e0f.

Declarations

Ethics approval and consent to participate

The “Therapeutic efficacy and mechanisms of MSC-derived MVs for RA-ILD in mice” study was approved by the Anima Ethical Committee of West China Hospital of Sichuan University (20240515007, 2024.05.15). The hUC-MSC study protocol of Kang Jing Co., Ltd. (Chengdu, CA) was approved by the Medical Ethics Committee of Wenjiang District People’s Hospital, Chengdu City (2024064, 2024.12.29), and the donors provided signed informed consent.

Consent for publication

All the authors have consented to publication.

Competing interests

The authors declare no competing interests.